RocketZephyr-3b / config.json
TitleOS's picture
The first attempt at merging stabilityai/stablelm-zephyr-3b and pansophic/rocket-3B with a linear merge method, 1.0 weight on zephyr and 0.3 on rocket.
fc9a196
raw
history blame
892 Bytes
{
"_name_or_path": "pansophic/rocket-3B",
"architectures": [
"StableLMEpochForCausalLM"
],
"auto_map": {
"AutoConfig": "stabilityai/stablelm-3b-4e1t--configuration_stablelm_epoch.StableLMEpochConfig",
"AutoModelForCausalLM": "stabilityai/stablelm-3b-4e1t--modeling_stablelm_epoch.StableLMEpochForCausalLM"
},
"bos_token_id": 0,
"eos_token_id": 50279,
"hidden_act": "silu",
"hidden_size": 2560,
"initializer_range": 0.02,
"intermediate_size": 6912,
"max_position_embeddings": 4096,
"model_type": "stablelm_epoch",
"norm_eps": 1e-05,
"num_attention_heads": 32,
"num_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"rope_pct": 0.25,
"rope_theta": 10000,
"rotary_scaling_factor": 1.0,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.35.2",
"use_cache": true,
"vocab_size": 50304
}